How Robot Mower Navigation Works

Strip away the marketing and every robot mower is solving the same two problems:

1. Where am I?
2. Where am I not allowed to go?

Every navigation technology on the market is an answer to one or both. Once you see them that way, the specification sheets become much easier to read.

The boundary wire: answering only the second question

The oldest approach ignores the first question entirely. A cable carrying a weak signal runs around the lawn; the mower detects the field and turns away when it reaches it.

The mower has no idea where it is. It simply knows when it has reached the edge. Between edges it drives more or less at random, which is why wire-based mowers can take a long time and cross the same ground repeatedly.

The strength of this approach is its indifference to everything else. No sky, no light, no landmarks — just a cable. In a heavily wooded garden this is still the most dependable technology available.

Satellite positioning: answering the first question, roughly

Adding a satellite receiver lets the mower estimate where it is. On its own, consumer-grade satellite positioning is accurate to a few metres, which is nowhere near enough to keep a mower off a flower bed.

That is why plain satellite positioning is rarely used alone. It typically supplements another system — helping the mower know roughly which part of the garden it is in.

RTK: answering the first question precisely

RTK adds a reference station that knows exactly where it is standing, so it can measure the current satellite error and broadcast a correction. That station is either an antenna fixed on your own property, or part of a permanent network reached over mobile data. The mower applies it and gets a position accurate to a few centimetres.

That accuracy is enough to answer both questions at once. A boundary becomes a line on a map rather than a physical cable, and the mower can cut in tidy parallel lines instead of wandering.

The limitation is inherent: it depends on seeing the sky. Under trees, beside tall buildings, in narrow passages, the correction degrades or is lost.

LiDAR: answering both questions from the garden itself

A laser scanner sweeps the surroundings and measures the time light takes to return, building a live map of walls, fences, trees and furniture.

This answers both questions without any reference to the sky. It also handles obstacles inherently: a system already mapping its surroundings sees the garden chair and drives around it.

Its weakness is the mirror image of RTK's. LiDAR needs features to navigate by. A vast, flat, empty lawn gives it relatively little to work with, and heavy rain or fog scatters the laser light.

Vision: answering both questions the way a person would

Cameras identify the lawn edge, obstacles, and sometimes the grass itself. Vision systems can recognise categories of thing — a hedgehog, a hosepipe, a shoe — rather than just detecting that something is there.

Performance depends on light. Deep shade, dappled sun, dusk and heavy rain are all harder than a bright even day.

Bump sensors and lift sensors: the safety net

Underneath all of this, essentially every mower still has physical sensors: a bumper that detects contact, and a lift sensor that stops the blades instantly if the machine is picked up or tipped.

These are not navigation. They are the last line of defence, and they matter most for safety.

Why combinations are winning

Each technology fails in a specific, well-understood way. Satellites fail under canopy. LiDAR struggles in featureless space. Vision struggles in poor light. Wires fail when cut.

Combining two whose weaknesses do not overlap produces something considerably more robust than either alone — which is why the better current models increasingly use satellite positioning in the open and on-board sensing everywhere else.

When comparing models, the useful question is not "which technology is best" but "which combination covers the specific weakness my garden will expose".